Where the Phosphorus in UK Sewage Comes From: A Per-Capita Budget and What It Means for Consents

Why the source budget matters

Phosphorus is the most common reason for water bodies in England failing good ecological status: around 55 percent of assessed rivers and 73 percent of assessed lakes miss the phosphorus standard, and sewage effluent supplies 60 to 80 percent of the phosphorus load reaching English rivers. Treatment works loads were cut by more than half between 1995 and 2010 and by about two thirds by 2020, at a capital cost of roughly £2.1 billion, yet tighter limits keep arriving: 0.5 mg/L annual averages are routine and 0.25 mg/L is now set at sensitive sites. Against that backdrop, knowing what arrives at the inlet and why is not academic. It fixes the removal percentage a process must deliver, the coagulant mass it will consume, the sludge it will make and the case a discharger can put to the regulator about source control.

The per-capita budget

UKWIR-sponsored source apportionment by Comber and colleagues, published in 2013, remains the most complete UK mass balance. Converted to grams of phosphorus per person per day, the domestic terms are:

Two independent checks anchor the human term. Dietary intake from natural foods (1.32 g P/day), supplements (0.06) and additives (0.59) sums to 1.97 g P/day; subtracting the 0.55 g/day retained by the body gives 1.4 g/day excreted, exactly the urine plus faeces figure measured elsewhere. On the works side, a crude sewage concentration of 8.25 mg P/L at 250 L per person per day of total inflow gives 2.3 g/day, of which the domestic budget explains about 93 percent once losses to combined sewer overflows and misconnections are deducted. Trade effluent, road runoff (around 0.66 mg P/L, a tenth of sewage strength) and commercial premises therefore make up only a few percent in a typical catchment, although an individual dairy, abattoir or metal finisher can dominate locally.

Diet, food additives and excreta

Roughly 70 percent of domestic phosphorus passes through a person. Natural dietary phosphorus comes mostly from cereals, dairy and meat and is fairly constant across western populations at 1.1 to 1.5 g/day. Phosphate additives are the growing part: polyphosphates and orthophosphates used as emulsifiers, moisture retainers, cure colour developers and acidity regulators in processed meats, spreadable cheese, instant desserts and cola drinks. Comber estimated 0.59 g P/day from additives in the UK diet, and American dietary work puts the range at 0.5 to 1.0 g/day depending on how processed the diet is. Unlike detergent phosphate, additive phosphorus cannot be removed by a product regulation on the detergent aisle; it tracks food policy and eating habits, and designers should treat 1.3 to 1.5 g P/day of excreted phosphorus as a floor that will not fall.

Excreted phosphorus is almost entirely soluble orthophosphate in urine and a mix of orthophosphate and organically bound phosphorus in faeces. On arrival at the works, after hydrolysis in the sewer, typically 60 to 80 percent of influent total phosphorus is reactive orthophosphate, which is what chemical and biological removal act on; the remainder is particulate and partly removed with primary sludge.

Detergents after the bans

In the 1970s detergents supplied about half of sewage phosphorus. Voluntary reformulation to zeolite builders through the 1990s and 2000s cut the laundry share, and then regulation closed it: the UK Detergents Regulations 2010 prohibited domestic laundry detergents above 0.4 percent phosphorus by weight from 1 January 2015, and Regulation (EU) 259/2012, now retained in GB law, capped consumer laundry detergents at 0.5 g P per standard dose from 30 June 2013 and consumer automatic dishwasher detergents at 0.3 g P per standard dose from 1 January 2017. At 2013 the dishwasher term was the one still rising, with 96 percent of the market phosphate-based and only 40 percent of households owning a machine; the 2017 cap removed most of it.

Denmark offers a well-documented view of the downstream effect. Analysis of inlet data from 56 domestic-only works by DTU for the Danish EPA found the per-capita reference value stable at about 1.10 kg P/PE/year (3.0 g/day) from 1990 to 2007, then falling by 34 percent to 0.72 kg P/PE/year (1.97 g/day) in 2017, with no sign of having reached a floor. UK inlet data show the same trajectory, and a works designed on the 2.5 g P/day values common in older guidance will now see 1.7 to 2.0 g/day. Phosphonate chelants such as HEDP and DTPMP remain in most laundry products at around 2.5 percent by weight; they are poorly degraded but adsorb strongly to sludge, so they contribute to the works load but little to the effluent.

Orthophosphate dosed for plumbosolvency

About 91 percent of the UK population receives drinking water dosed with orthophosphoric acid or sodium phosphate to suppress lead dissolution, in response to the 10 µg/L lead standard applied from December 2013. Company target concentrations range from about 0.7 to 1.9 mg P/L, with a supply-weighted mean of 0.91 mg P/L. Because the dosed phosphate equilibrates with the pipework, almost all of it reaches the tap and then the drain. At 150 L per person per day of domestic use the term is 0.13 g P/day; counting all dosed water in the catchment, including commercial and industrial supply, it is nearer 11 percent of the total works load.

Because it is dosed at 1 mg/L or so, this source alone would exceed a 0.5 mg/L consent at a works with no phosphorus removal, and it is the only term the water industry itself controls. Reducing it means lead pipe replacement, estimated at £890 million for one company serving seven million people, so it is unlikely to fall quickly. Where a works discharges to a supply catchment the phosphorus is partly recycled: abstraction, dosing, use, treatment and discharge in a loop.

Worked example: a 10,000 PE works

Take a works serving 10,000 population equivalent with a dry weather flow of 2,000 m³/day (200 L/PE/day including infiltration) and a current per-capita load of 1.8 g P/day.

Chemical removal with ferric at a molar Fe:P ratio of 1.5 for the 0.5 mg/L case consumes 1.5 × (55.85 / 30.97) = 2.7 kg Fe per kg P removed, so about 17 kg P/day removed needs 46 kg Fe/day, or around 115 kg/day of 40 percent ferric chloride solution, and generates roughly 60 kg/day of extra dry solids as ferric phosphate and hydroxide. Run the same calculation on the 2.5 g P/day of older design guidance and the chemical and sludge figures rise by 40 percent, which is why re-basing the load on current inlet data before an upgrade is usually the cheapest design decision on the project. Conversely, a works whose inlet sampling shows 2.5 g P/day or more should look for a trade input, septic tank imports or a dairy before concluding that the population is unusual. Sizing of the dosing and settlement stages is covered in the chemical phosphorus removal guide, and the ultra-low tertiary case in the tertiary phosphorus removal article.

Consent and policy implications

• Load-based consents and nutrient neutrality. Where permits or Habitats Regulations assessments are expressed as kg P/year, the per-capita figure sets the budget for new housing. Using 1.7 to 2.0 g P/day rather than legacy values of 2.5 or more can reduce the offset a developer must buy by a fifth or more, and the regulator will accept measured inlet data from comparable works as evidence. The compliance mechanics are in the nutrient neutrality guide.
• Removal percentage. Falling influent concentration does not loosen a concentration consent; it raises the required removal percentage slightly less than proportionally, and it lowers the orthophosphate residual that chemical dosing can reach, since the equilibrium residual depends on the dose ratio rather than the starting concentration.
• Recovery potential. A falling load also shrinks the struvite or sludge-ash phosphorus available for recovery, which bears on business cases discussed in the struvite recovery article.
• Preventable fraction. Comber estimated that tap water dosing, food additives, food waste, detergents and personal care products together make up about half of the load to works; of these only detergents have so far been regulated. Source control of additives would need food policy, not water policy, and the industry argument that further gains at the works are disproportionate increasingly rests on this point.
• Small and private works. Septic tanks and package plants are small nationally but can dominate headwater catchments, and imports of tankered septic waste can raise a receiving works load far above the domestic budget.

Measuring the load properly

Per-capita loads should be derived from flow-proportional 24-hour composites over at least a fortnight in dry weather, with the population served taken from billing records rather than design figures, and with total and reactive phosphorus analysed separately so the orthophosphate fraction is known. Grab samples in the morning peak overstate the load; samples taken during rain understate it because of dilution and because a share is lost at overflows. A works that logs inlet phosphorus continuously with an orthophosphate analyser can also see the weekly dishwasher and laundry pattern directly, which is a useful check on whether residual detergent sources remain in the catchment.